The prediction model of genetic and risk factors to hemorrhoids

Abstract

Hemorrhoids is the most common anorectal disease. At present, the research on hemorrhoids has focused on the external environment and behavioral factors, while the research on genetic factors is rare. This study used mixed methods to establish a preliminary hemorrhoid risk prediction model with genetic and risk factors, providing clues and theoretical basis for further research on hemorrhoid prevention. The research objectives of this study include the following four aspects: (1)To explore the risk factors of hemorrhoids in Guangxi population. (2)To explore the causal association of body mass index and hemorrhoids. (3) To explore the association of FOXC2 polymorphism and its interaction with other risk factors with the susceptibility to hemorrhoids in Guangxi population. (4)To establish a prediction model for hemorrhoids.

This study can be divided into two phases. The first phase includes two parts. In the first part, a case-control study was used to analyze the relationship between the indicators in the questionnaire survey and the rs34221221 polymorphism of FOXC2 gene and hemorrhoids. The second part is to analyze the causal relationship between BMI and hemorrhoids using Mendelian randomization method.

In the case-control study, 200 patients with hemorrhoids and 200 patients without hemorrhoids were included according to the inclusion and exclusion criteria. A self-made questionnaire was used to collect the general information, environmental and behavioral factors that may be related to hemorrhoids, and the venous blood of patients was collected for single nucleotide polymorphism detection. T-test was used to analyze the quantitative data of the two groups, Chi-square test was used to analyze the classification data, and logistic regression was used to analyze the correlation between variables and hemorrhoids. The results showed that age, high education level, constipation and chronic gastritis will increase the risk of hemorrhoids (Adjusted OR=1.03, 95% CI: 1.01-1.05; Adjusted OR=2.02, 95% CI: 1.16-3.52; Adjusted OR=2.70, 95% CI: 1. 50-4. 87; Adjusted OR=1.96, 95% CI: 1.14-3.36), while bland dietary taste and CT genotype could reduce the risk of hemorrhoids (Adjusted OR=0.50, 95% CI: 0.28-0.89, Adjusted OR=0.642, 95% CI: 0.424-0.972).

In the second part, using BMI as the exposure factor and hemorrhoids as the outcome, a two sample Mendelian randomized analysis was performed. A total of 68 SNPs of BMI in east asian population were selected as genetic instrumental variables, and total of 67 SNPs in south asian population were included. The causal relationship between BMI and hemorrhoids was analyzed using IVW, MR- Egger and weighted median models. No causal effect was found between BMI and hemorrhoids. In East Asian population, in IVW model, B=1.790, 95% CI: 0.628-5.107, P=0.276; In MR-Eegger model, B=0.316, 95% CI :0.013-7.908, P=0.486; In the WME model, B=4.018, 95% CI: 0.816-9.794, P=0.087. In the South Asian population, in the IVW model, B=0.753, 95% CI 0.417-1.362, P=0.349; In MR-Egger model, B=0.642, 95% CI : 0.144-2.862, P=0.563; In WME model, B=0.369, 95% CI: 0.146-0.993, P=0.035. Cochran's Q test results showed that there was no heterogeneity in instrumental variables of MR analysis (P>0.05), and there was no pleiotropy in MR-Egger based on intercept (P>0.05). MR-PRESSO results showed that there was no outlier (Gglobal>0.05).

The second phase is to use the variables related to hemorrhoids discovered in the first phase as independent variables and hemorrhoids as the dependent variable for logistic regression analysis. The regression coefficients of the obtained independent variables are used as weights to establish a hemorrhoid risk prediction model(Logit(P)=-1.4+0.03X1+0.79X2-0.71X3+1.02X4+0.65X5-0.45X6, where X1 represents age, X2 represents education level, X3 represents dietary taste, X4 represents constipation, X5 represents chronic gastritis, and X6 represents the genotype of the rs34221221 locus of the FOXC2 gene). The sensitivity of this model is 0.561, the specificity is 0.645, the cut-off value is 0.6, the accuracy of predicting the incidence of hemorrhoids is 76%, the accuracy of predicting the incidence of non hemorrhoids is 90%, and the accuracy of predicting the incidence of hemorrhoids in the entire group is 87.8%.

In conclusion, This study found that age, education level, constipation, and chronic gastritis can increase the risk of hemorrhoids, while a bland diet can reduce the risk of hemorrhoids; the rs34221221 polymorphism of the FOXC2 gene is associated with susceptibility to hemorrhoids, people with CT genotype are less susceptible to hemorrhoids than those with CC genotype. The final established model is Logit(P)=-1.4+0.03X1+0.79X2-0.71X3+1.02X4+0.65X5-0.45X6, when the predicted value is greater than 0.6, the predicted result indicates the possibility of developing hemorrhoids, and if it is less than 0.6, it indicates the possibility of not developing hemorrhoids.

-